Methods and systems for monitoring band structure characteristics and predicting electrical characteristics of a sample early in a semiconductor manufacturing process flow are presented herein. High throughput spectrometers generate spectral response data from semiconductor wafers. In one example, the measured optical dispersion is characterized by a Gaussian oscillator, continuous Cody-Lorentz model. The measurement results are used to monitor band structure characteristics, including band gap and defects such as charge trapping centers, exciton states, and phonon modes in high-K dielectric layers and embedded nanostructures. The Gaussian oscillator, continuous Cody-Lorentz model can be generalized to include any number of defect levels. In addition, the shapes of absorption defect peaks may be represented by Lorentz functions, Gaussian functions, or both. These models quickly and accurately represent experimental results in a physically meaningful manner. The model parameter values can be subsequently used to gain insight and control over a manufacturing process.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A system comprising: an illuminator configured to provide an amount of illumination to an unfinished, multi-layer semiconductor wafer across a spectral range; a spectrometer configured to collect an amount of light from the unfinished, multi-layer semiconductor wafer in response to the illumination provided by the illuminator and generate an amount of data indicative of a spectral response of the unfinished, multi-layer semiconductor wafer; and one or more computer systems configured to: receive the spectral response of the unfinished, multi-layer semiconductor wafer across the spectral range; determine a plurality of parameter values of an optical dispersion model of one or more layers of the multi-layer semiconductor wafer based at least in part on the spectral response, wherein the optical dispersion model includes a continuous Cody-Lorentz model having a first derivative function that is continuous at the Urbach transition energy of the continuous Cody-Lorentz model and at least one unbounded Gaussian oscillator function; and store the plurality of parameter values of the optical dispersion model in a memory.
A system for analyzing semiconductor wafers comprises an illuminator that shines light across a range of colors onto a multi-layer wafer during manufacturing. A spectrometer measures the light reflected from the wafer and creates spectral response data. A computer then receives this data and determines parameters for a model that describes how light interacts with the wafer's layers. This model uses a "continuous Cody-Lorentz" function (with a smooth transition at a specific energy level) and at least one "unbounded Gaussian oscillator" to characterize the wafer's optical properties. The computer stores these model parameters in memory.
2. The system of claim 1 , wherein the optical dispersion model is sensitive to a band gap of a layer of the unfinished, multi-layer semiconductor wafer.
The system described previously, which analyzes semiconductor wafers using light and a computer model, uses an optical dispersion model (containing Cody-Lorentz and Gaussian functions) that is sensitive to the band gap of a layer within the unfinished, multi-layer semiconductor wafer. The determined parameter values provide information about the band gap.
3. The system of claim 1 , wherein the one or more computer systems are further configured to: determine a band structure characteristic indicative of an electrical performance of a first layer of the multi-layer semiconductor wafer based at least in part on parameter values of the optical dispersion model of the multi-layer semiconductor wafer.
The system for analyzing semiconductor wafers described previously, uses its computer system to calculate a "band structure characteristic" that indicates the electrical performance of a layer in the wafer. This calculation relies on the parameter values derived from the optical dispersion model (Cody-Lorentz and Gaussian functions) of the multi-layer semiconductor wafer. The band structure characteristic reveals information about the electrical properties of the semiconductor wafer.
4. The system of claim 3 , wherein the electrical performance of the multi-layer semiconductor wafer is any of an equivalent oxide thickness (EOT), a leakage current, a threshold voltage, and a breakdown voltage.
In the system that analyzes semiconductor wafers and calculates electrical performance, the electrical performance metric can be any of these parameters: equivalent oxide thickness (EOT), leakage current, threshold voltage, or breakdown voltage. These metrics, derived from the optical model parameter values, provide indicators of the semiconductor wafer quality.
5. The system of claim 3 , wherein the one or more computer systems are further configured to: control a process of manufacture of the unfinished, multi-layer semiconductor wafer based at least in part on the band structure characteristic.
The system described previously, which analyzes wafers and determines their electrical performance based on an optical model (Cody-Lorentz and Gaussian functions), further uses the calculated "band structure characteristic" to control the wafer manufacturing process. The system adjusts manufacturing parameters based on the analysis results to improve wafer quality.
6. The system of claim 1 , wherein one or more layers of the multi-layer semiconductor wafer include at least one nanostructure.
The system, previously described, analyzing the unfinished, multi-layer semiconductor wafer which uses an optical model, contains one or more layers that include at least one nanostructure. These nanostructures could influence the optical properties measured by the system.
7. The system of claim 6 , wherein the at least one nanostructure is any of a plurality of quantum dots, a plurality of nanowires, and a plurality of quantum wells.
In the system that analyzes semiconductor wafers with layers containing nanostructures, the nanostructures can be quantum dots, nanowires, or quantum wells. These structures are embedded within the layers of the wafer and are analyzed by the optical measurement system.
8. The system of claim 1 , wherein a first layer of the multi-layer semiconductor wafer is an electrically insulative layer disposed above a semiconductor substrate.
In the wafer analysis system, a layer of the unfinished, multi-layer semiconductor wafer is an electrically insulative layer that sits on top of a semiconductor substrate. The system analyzes this insulating layer and its properties using the optical dispersion model.
9. The system of claim 8 , wherein the first layer includes an intermediate layer between the semiconductor substrate and the electrically insulative layer.
In the semiconductor wafer analysis system which analyzes the insulative layer disposed above a semiconductor substrate, an intermediate layer exists between the substrate and the insulating layer. The system's optical analysis also accounts for the presence and properties of this intermediate layer.
10. The system of claim 1 , wherein the illuminator and spectrometer are configured as any of an ellipsometer and a reflectometer.
In the system for analyzing semiconductor wafers, the illuminator (light source) and spectrometer (light detector) are configured as either an ellipsometer or a reflectometer. These instruments measure changes in the light's polarization or intensity upon reflection, respectively, and are used to generate the spectral response data.
11. A method comprising: receiving a spectral response of an unfinished, multi-layer semiconductor wafer across a spectral range; determining a plurality of parameter values of an optical dispersion model of one or more layers of the multi-layer semiconductor wafer based at least in part on the spectral response, wherein the optical dispersion model includes a continuous Cody-Lorentz model having a first derivative function that is continuous at the Urbach transition energy of the continuous Cody-Lorentz model and at least one unbounded Gaussian oscillator function; and storing the plurality of parameter values of the optical dispersion model in a memory.
A method for analyzing semiconductor wafers involves receiving spectral response data across a range of colors from an unfinished, multi-layer wafer. The method determines parameters for an optical dispersion model of one or more layers within the wafer, based on the spectral response. The model uses a "continuous Cody-Lorentz" function (with a smooth transition at a specific energy level) and at least one "unbounded Gaussian oscillator" to characterize the wafer's optical properties. The resulting parameter values are then stored in memory.
12. The method of claim 11 , wherein the optical dispersion model is sensitive to a band gap of a layer of the unfinished, multi-layer semiconductor wafer.
The method described previously, analyzing semiconductor wafers and determining the optical model parameters, uses an optical dispersion model (containing Cody-Lorentz and Gaussian functions) that is sensitive to the band gap of a layer within the unfinished, multi-layer semiconductor wafer. The determined parameter values provide information about the band gap.
13. The method of claim 11 , further comprising: determining a band structure characteristic indicative of an electrical performance of a first layer of the multi-layer semiconductor wafer based at least in part on parameter values of the optical dispersion model of the multi-layer semiconductor wafer.
The method for analyzing semiconductor wafers, which involves receiving spectral data and creating an optical model (Cody-Lorentz and Gaussian functions), further includes calculating a "band structure characteristic" that indicates the electrical performance of a layer in the wafer. This calculation relies on the parameter values derived from the optical dispersion model of the multi-layer semiconductor wafer.
14. The method of claim 13 , wherein the electrical performance of the multi-layer semiconductor wafer is any of an equivalent oxide thickness (EOT), a leakage current, a threshold voltage, and a breakdown voltage.
In the method that analyzes semiconductor wafers and calculates electrical performance, the electrical performance can be measured by metrics such as equivalent oxide thickness (EOT), leakage current, threshold voltage, or breakdown voltage. These metrics, derived from the optical model parameter values, provide indicators of the semiconductor wafer quality.
15. The method of claim 13 , further comprising: controlling a process of manufacture of the unfinished, multi-layer semiconductor wafer based at least in part on the band structure characteristic.
The method described previously, analyzing wafers and determining their electrical performance based on an optical model (Cody-Lorentz and Gaussian functions), further uses the calculated "band structure characteristic" to control the wafer manufacturing process. The method adjusts manufacturing parameters based on the analysis results to improve wafer quality.
16. A non-transitory, computer-readable medium, comprising: code for causing a computer to receive a spectral response of an unfinished, multi-layer semiconductor wafer across a spectral range; code for causing the computer to determine a plurality of parameter values of an optical dispersion model of one or more layers of the multi-layer semiconductor wafer based at least in part on the spectral response, wherein the optical dispersion model includes a continuous Cody-Lorentz model having a first derivative function that is continuous at the Urbach transition energy of the continuous Cody-Lorentz model and at least one unbounded Gaussian oscillator function; and code for causing the computer to store the plurality of parameter values of the optical dispersion model in a memory.
A computer-readable medium stores instructions to analyze semiconductor wafers by receiving spectral response data across a range of colors from an unfinished, multi-layer wafer. The instructions cause the computer to determine parameters for an optical dispersion model of one or more layers within the wafer, based on the spectral response. The model uses a "continuous Cody-Lorentz" function (with a smooth transition at a specific energy level) and at least one "unbounded Gaussian oscillator" to characterize the wafer's optical properties. The parameter values are then stored in memory.
17. The non-transitory, computer-readable medium of claim 16 , wherein the optical dispersion model is sensitive to a band gap of a layer of the unfinished, multi-layer semiconductor wafer.
The computer-readable medium, described previously, that analyzes semiconductor wafers and determines the optical model parameters, uses an optical dispersion model (containing Cody-Lorentz and Gaussian functions) that is sensitive to the band gap of a layer within the unfinished, multi-layer semiconductor wafer. The determined parameter values provide information about the band gap.
18. The non-transitory, computer-readable medium of claim 16 , further comprising: code for causing the computer to determine a band structure characteristic indicative of an electrical performance of a first layer of the multi-layer semiconductor wafer based at least in part on parameter values of the optical dispersion model of the multi-layer semiconductor wafer.
The computer-readable medium for analyzing semiconductor wafers, which involves receiving spectral data and creating an optical model (Cody-Lorentz and Gaussian functions), further includes instructions that cause the computer to calculate a "band structure characteristic" that indicates the electrical performance of a layer in the wafer. This calculation relies on the parameter values derived from the optical dispersion model of the multi-layer semiconductor wafer.
19. The non-transitory, computer-readable medium of claim 18 , wherein the electrical performance of the multi-layer semiconductor wafer is any of an equivalent oxide thickness (EOT), a leakage current, a threshold voltage, and a breakdown voltage.
In the computer-readable medium that analyzes semiconductor wafers and calculates electrical performance, the electrical performance can be measured by metrics such as equivalent oxide thickness (EOT), leakage current, threshold voltage, or breakdown voltage. These metrics, derived from the optical model parameter values, provide indicators of the semiconductor wafer quality.
20. The non-transitory, computer-readable medium of claim 18 , further comprising: code for causing the computer to control a process of manufacture of the unfinished, multi-layer semiconductor wafer based at least in part on the band structure characteristic.
The computer-readable medium, described previously, that analyzes wafers and determines their electrical performance based on an optical model (Cody-Lorentz and Gaussian functions), further uses the calculated "band structure characteristic" to control the wafer manufacturing process. The medium stores instructions that cause the computer to adjust manufacturing parameters based on the analysis results to improve wafer quality.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
May 19, 2016
May 30, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.